Magnetic resonance imaging (MRI) has become the gold-standard technique for non-invasive in-vivo visualization of clinical anatomy. However this cannot be said for MRI's application to microscopic visualization of biological tissues and single cells. Standard MRI techniques are limited in resolving power by a variety of factors, most importantly diffusion, magnetic susceptibility differences, and intrinsic NMR linewidth. It is generally agreed that diffusion is the most egregious of the factors to limiting the resolution of MRI in biological samples. Most calculations place a practical limit on standard MRI techniques on the order of ( 10 micromole)3 in these samples. We propose a MRI microscope which will resolve cells, both singly or within tissues, as well as sub-cellular structures. This is based on the DESIRE (Diffusion-Enhanced Signal Intensity and REsolution) experiment described by Lauterbur and colleagues which uses the microscopic effects of diffusion to amplify the MRI signal from volumes smaller than those possible with standard MRI techniques. In the DESIRE technique, diffusion of spins through a small region to their surroundings stores magnetization outside the voxel and uses that signal to enhance the signal-to-noise ratio from the small volume. Calculations show that for favorable conditions signal amplification as high as 6 orders of magnitude may be possible. Small lithographically- constructed MRI surface coils will be used to implement this experiment utilizing positioning technology developed in scanning tunnelling microscopy. The goal of this proposal is the construction and testing of a DESIRE MRI microscope capable of about (1 micromole)3 resolution on a biological sample.